Gemini 2.5 Pro Preview 05-06 is a near-frontier model from Google, sitting just below the top intelligence tier with an index of 29.5. Its LiveCodeBench score of 0.770 is among the strongest in the broader landscape, and it supports tool use, function calling, and vision across text, image, audio, video, and file inputs. The 1M token context window is a standout feature for long-document and large-codebase workflows.

For businesses, this model is a strong fit for software engineering, autonomous coding agents, long-document analysis, and multimodal content workflows. The breadth of supported modalities means teams can consolidate multiple task types on a single model. Its intelligence index of 29.5 places it in the strong-but-not-frontier tier, and it is superseded by later Gemini releases in the top intelligence rankings.

At $1.25 input and $10.00 output, output pricing is high and will add up quickly on token-heavy tasks. Teams running high-volume pipelines should model costs carefully. For lower-volume, high-value tasks such as complex code review, contract analysis, or multimodal document processing, the capability breadth justifies the cost.

Assessed July 10, 2026

Rankings consider pricing, capabilities, benchmarks, and real-world applicability and are refreshed as new models launch. Feedback?